The “short or long but never midrange” theory of NBA shooting has been popular in analytics since at least the early 2000s. Statistically, 3-point shots have long been more efficient than long 2-point shots: The reward — an extra point — has outweighed a slightly lower completion percentage. So, the theory goes, teams should eschew midrange shots as much as possible and either take the ball to the basket or shoot from beyond the arc.
Whether or not NBA players and coaches have had this theory in mind, 3-point attempts have steadily risen since their introduction in the 1979-80 season. In “Daryl Morey’s D-League Plan to Do Away With Midrange Shots,” Grantland’s Kirk Goldsberry looked at the Rio Grande Valley Vipers, the Houston Rockets D-League team that has taken the inside/outside strategy to the extreme — a staggering 47 percent of players’ shots come from the 3-point range. (If the NBA keeps increasing its long-distance shooting at the pace it has in the past 25 years, it should match the Vipers in approximately 2054.)
Although the Vipers have had good results so far, Goldsberry warned against reading too much into the experiment and assuming the same strategy would work in the NBA:
One troubling issue with the Vipers’ strategy is that they are significantly reducing the size of the scoring area. Let’s go back to the 1,300-square-foot apartment metaphor — the Vipers hang out only in the kitchen and bedroom. They never use the living room. By forgoing midrange shots altogether, they are telling opponents that they don’t plan on using a giant swath of tactical space relatively close to the basket. As NBA defenses evolve, the smartest ones are already finding ways to counter the smart-guy shot-selection strategy. The Pacers, who have the league’s best defense, have a reputation for vigorously defending the restricted area and the 3-point area, while being less protective of the midrange. They are the Vipers’ antivenom. And if defenses are able to take away those “efficient” areas, you better have alternative ways to generate offense.
Goldsberry was describing an important limitation in the 3s-and-layups strategy: 3-point shots will only be more efficient than midrange shots so long as defenses fail to devote enough resources to defending them. Unless the best-defended 3-point shot is more efficient than a wide-open midrange shot (it’s not), saying 3-point shots are better than midrange shots is wrong. What we really mean is 3-pointers are suboptimally defended. As unwitting game-theory expert Stan Van Gundy would say, “Sometimes you’re going to get the shots the defense allows.”
But there’s an even bigger issue with the back-of-the-envelope efficiency calculation at the heart of the pro-3 argument: The fact that 3-pointers are more efficient than midrange shots overall doesn’t mean the marginal 3-pointer is necessarily more valuable than the marginal 2-pointer.
If you want to replace a midrange attempt with a 3-pointer, you will normally be replacing a bad 2-pointer with a bad 3-pointer. There’s no intrinsic reason those marginal shots should have the same relative efficiency as better shots of the same type. In fact, there are plausible reasons to think they wouldn’t. Average 3-point shot efficiency may be skewed by the extremely high efficiency of wide-open looks, which players probably weren’t passing up in the first place.
This doesn’t mean that NBA teams shouldn’t be taking more 3s. I’d guess we probably haven’t reached that point yet.